IEC 61620 Insulating liquids - Determination of the dielectric dissipation factor by measurement of the conductance and capacitance - Test method
8 Procedure
In order to get a significant measurement of dissipation factor tan δ, it is essential to follow precise rules concerning
- the careful cleaning of the test cell;
- the careful filling of the test cell and handling of the liquid samples and of the test cell itself.

8.1 Cleaning of the test cell
8.1.1 Operating procedure
According to the state of cleanliness of the test cell and the level of conductivity of the liquid to be measured, the cleaning procedure of the test cell can be more or less sophisticated and take more or less time.

If the cleanliness of the test cell is unknown or if there is any doubt, a cleaning procedure shall be applied.

Many different types of cleaning procedure can be used provided they have proved to be efficient.

In Annex A, a reference procedure is given. It shall be used in case of dispute between two laboratories. In Annex B, an appropriate simplified cleaning procedure is given as an example.

NOTE For routine testing and when a number of samples of the same type of unused liquid are to be tested consecutively, the same test cell may be used without intermediate cleaning, provided that the value of the property for the sample previously tested is better than the specified value. If such is not the case, the test cell must be cleaned before being used for further tests.

8.1.2 Checking the cleanliness of the empty cell To obtain a significant measurement it is necessary that the electrical losses of the empty cell be much lower than those of the liquid to be measured.

NOTE Nevertheless, the walls of the vessel and the electrodes may retain impurities liable to dissolve eventually in the liquid.

8.1.3 Checking the cleanliness of the filled cell for measurement at room temperature
If the test cell is perfectly clean and if the temperature of the liquid is constant, σ and tan δ are independent of time. The measurement can thus be taken as soon as it is practically possible. In fact, this can be done within 1 min. Moreover, one measurement on a single sample is enough to obtain the right value.

It may happen that, at constant temperature, the conductivity σ (or tan δ) increases or decreases with time, but not by more than 2 % at two minutes after filling. In this case the test cell is considered sufficiently clean and the first measured value, i.e. at one minute or less after filling, may be recorded.

If not, it is recommended, after having cleaned the test cell again, to take a second sample of the same liquid and to carry out a second measurement. The lowest of the two values shall be retained as already recommended in IEC 60247.

8.1.4 Checking the filled cell for measurement at temperature above room temperature
Before making any measurement at high temperature, ensure that the temperature of the liquid throughout the cell is constant. Except for cases where the test cell is perfectly clean, the result of the measurement will depend on the way in which the test cell and the liquid are brought to high temperature.

If the test cell is perfectly clean and if the temperature of the liquid is constant, the conductivity σ and tan δ are independent of time. The measurement can thus be taken immediately. Practically, the measurement can be taken as soon as the temperature can be considered constant. Moreover, one measurement on a single sample is enough to obtain the correct value.

It may happen that, even though the liquid in the test cell is at constant temperature, the conductivity σ (or tan δ) increases or decreases with time. This can be due to different causes: e.g. heating at elevated temperatures may alter the composition of certain insulating liquids, or modify the moisture content of particles. In practice, temperature is not perfectly constant and its variations naturally influence the possible variations of σ (or tan δ). The conductivity σ (or tan δ) of liquids more or less varies with the temperature according to the nature of the liquid, typically up to 5 % per degree Celsius. Therefore, the origin of the variations of σ (or tan δ) can only be determined if the temperature fluctuations are small enough. Then, if the drift of σ (or tan δ) is less than 2 % over 2 min, the test cell is considered sufficiently clean, and the value measured one minute or less after temperature is considered as constant can be recorded.

If the test cell is not perfectly clean, the heating time will influence the values measured, even the first one, because the impurities coming from the test cell dissolve into the liquid. The first value measured shall therefore be rejected, and the test cell cleaned again.

8.2 Precautions to be taken when filling the test cell
When filling the cell with the sample, ensure that the environmental atmosphere is, as far as possible, free of any vapours or gases liable to dissolve in the liquid.

The electrodes shall be entirely immersed in the liquid.

NOTE When not in use, the cell must be stored in a desiccator.

8.3 Test temperature
Measurement of the conductivity and of the dissipation factor of a liquid can be performed at any temperature.

Ambient temperature is recommended for its simplicity of operation and time-saving qualities. The ambient temperature being essentially variable, a fixed value shall be agreed (for instance 25 °C +/- 1 °C).

There is nothing to prevent carrying out the test at a higher test temperature (for instance 40 °C +/- 1 °C, 90 °C +/- 1 °C or more).

8.4 Heating methods
To perform a measurement at high temperature several heating methods can be used. The time required to reach the test temperature will depend on the heating method and may typically vary from 10 min to 60 min. If the test cell is not perfectly clean, the increase in conductivity due to the progressive dissolution of impurities will depend on the duration of the heating period, and the measured conductivity will therefore depend on the method of heating.

Thus it is recommended to heat the test cell as fast as possible.

A suitable way to achieve this is to heat separately the test cell and the liquid contained in a clean vessel. Another method consists in rapidly heating the liquid in the test cell itself.

NOTE Rapidly heating the liquid and cell together may result in a significant thermal gradient. Before using such a method the uniformity of temperature should be verified for type of fluid being tested.

8.5 Measurement
Fill the cell with the sample avoiding any pollution of the liquid or of the cell (see 8.2).

Check cleanliness according to 8.1.3 and 8.1.4. If the cell is sufficiently clean (see 8.1), note the values of G and C.